Dynamic evaluation of unruptured intracranial aneurysms by 4D-CT angiography: comparison with digital subtraction angiography (DSA) and surgical findings

In this study, 4D-CTA was used prospectively to evaluate patients with UIAs and referenced to DSA, and in cases suspected of UIA instability, compared to the surgical presentation of the aneurysm. The overarching experimental conclusion was that 4D-CTA was reliable for the localization and anatomical characterization of UIAs with excellent agreement with DSA results and surgical findings. Importantly, high-resolution temporal acquisitions of UIA pulsation helped to recognize unstable smaller lesions that might be overlooked with static DSA imaging. Advantages of 4D-CTA over DSA included being a non-invasive and lower-cost technique that offered rapid multi-vessels, multi-angle imaging significantly lower effective radiation dose exposure [16, 17].

DSA is the current “gold standard” for the diagnosis of intracranial aneurysms and surgical strategy guide because of its ability to elucidate the UIA pedicle and origin as well as its spatial relationship with nearby vascular branches. The inherent drawbacks of DSA remain potential procedural complications and high costs associated with the invasive intravascular intervention [18]. Magnetic resonance angiography (MRA) is also challenging the routine use of DSA. The diagnostic advantages of MRA for UIA include its non-invasive nature, avoidance of contrast agents, avoidance of radiation, better soft tissue definition, and multi-directional reconstruction and display to enhance assessment of anatomical detail [19, 20]. UIA greater than 3 mm in diameter can be resolved by MRA with an accurate rate of 67–86% [21, 22].

3D-CTA has also evolved to challenge DSA as the reference standard. 3D-CTA is very rapid, low cost, and minimally invasive but does require contrast agents [19]. Improvements for single-slice spiral CT to 16 multi-slice CT have enhanced the detection of cerebral vascular branch vessels to 3–4 branches and increased UIA diagnostic sensitivity to aneurysms larger than 3 mm, similar to MRA [23]. Further development of 64-multislice spiral CT allowed the use of isotropic voxels, such that the reconstructed image achieves the same resolution within or through plane, and in any oblique cross-section. The improved angiography images were further refined through the continual development of more powerful reconstruction and post-processing software [24], which allowed 3D-CTA can make to achieve the same sensitivity, specificity, accuracy, and precision for noninvasive UIA evaluations as DSA [25].

Although aneurysms less than 3 mm have been considered to be the lower limit of detection for 3D-CTA, in the present study, using 320-multi-slice spiral CT, all aneurysms were well recognized on the CTA image including those less than 3 mm [26]. In general, many studies have reported that artifacts, the amount of contrast agent, and CT scan parameters greatly influenced and could detract from CT image quality. However, such artifacts have been minimized with new-generation scanners, such as in the present study, which employed advanced wide-detector CT scanning technology, which facilitated image post-processing results, morphological detail and quantitative measurements [27]. Moreover, the recent implementation of iterative reconstruction algorithms (i.e., maximum likelihood approaches) rather than traditional filtered back projection methods, as used in the present study can further improve image quality and reduce patient radiation exposure, even beyond the significant benefits achieved in this study over DSA [28]. Encouraging, in our experience, CTA has excellent consistency with DSA for the assessment of intracranial unruptured aneurysms. We found that CTA has a high detection rate in aneurysms, which may be due to the removal of bone structure interference by CTA, leading to a higher vascular resolution and increased detection rate of intracranial aneurysms.

Although noninvasive MRA and 3D-CTA enhance the details of UIA and surrounding anatomy, these methods lack the temporal resolution to characterize the dynamics of the aneurysms [29]. Aneurysm features are not always static within the cardiac cycle. They can undergo continuous morphological and volumetric changes that manifest as abnormal pulsation [30], as confirmed in the present study. Large UIA is more likely to rupture than smaller counterparts [31]. However, small aneurysms are also at risk of rupture [32], and discrimination of a small stable UIA versus one at higher risk of rupture is critical to patient risk-benefit stratification for therapeutic intervention. Increasingly, the use of static morphological parameters is questioned [33], with the adjunctive presence of UIA pulsation perhaps being critical to rupture risk decisions, particularly for small UIA [34]. However, while UIA pulsation is suggested to be important in the risk assessment of UIA, the requirements for CT imaging equipment capable of evaluating pulsation have been a practical clinical barrier [35]. While the development of 64-multi-slice spiral CT has found success in overcoming large-scale motions of the heart and the great vessels, requisite long scan time, small z-axis coverage, and low temporal resolution preclude capturing the needed CT scan data within a single cardiac cycle, which is required to observe the minute pulsation of UIAs [10].

In the present study using 320-multi-slice CT to perform 4D-CTA imaging, the 16 cm scan range permitted the field-of-view (FOV) to encompass the entire brain in one rotation of the X-ray tubes without table repositioning of the patient. The improved image quality and temporal resolution were substantial, which was reflected in the definitive detection of abnormal aneurysm pulsation that was subsequently confirmed intraoperatively [36]. In our study, all large cerebral aneurysms in the 4D-CTA imaging showed a significant pulsation, the patients received surgical treatment in time, successfully carried out an aneurysm clamping, effectively prevented and avoided the rupture of the aneurysms, greatly improved the patient’s clinical prognosis. The results of the operation confirmed that the abnormal pulsation site observed in 4D-CTA was corresponding to the weak area of the aneurysmal wall in surgery (as shown in Figs. 3 and 4). In addition, in this study, some of the relatively small cerebral aneurysms appeared stable, but in the 4D-CTA imaging, there also was an abnormal pulsation, the majority of patients actively engaged in craniotomy surgery or interventional embolization treatment, achieved good clinical results and avoided the occurrence of bad cerebrovascular events. Thus, Small and large aneurysms with foreboding pulsations were corroborated to be dark-reddish, thin-wall, high-risk lesions at surgery (as shown in Figs. 5 and 6). Aneurysm irregularity may reflect the focal weakness of the wall of the aneurysm, which is related to degeneration. These changes may lead to a variety of aneurysm wall remodeling, including endothelial cell damage, smooth muscle cell migration and inflammatory cell infiltration. Irregular pulsation will more likely develop, but underlying mechanisms require further studies [37].

Unfortunately, for one patient with abnormal aneurysmal pulsation detected by 4D-CTA, surgical intervention was deferred. However, 36 days later interventional embolization of a subarachnoid hemorrhage was required when the patient presented severe neurological symptoms (as shown in Fig. 7) was performed. It indicated that the pulsation of the aneurysm was a predictor of the risk of rupture of an aneurysm. In practice, when an abnormal pulsation was found in an unruptured intracranial aneurysm, it indicated a higher risk of rupture, suggesting that the neurosurgeon should take treatment as early as possible to avoid sudden rupture during follow-up. In contrast, the patients without UIA pulsations observed by 4D-CTA were discharged without intervention, and clinical neurological evidence of aneurysm ruptures was reported at the standard regular follow-up contact. We boldly believed that these aneurysms were in a relatively stable state, and currently, there is no occurrence of aneurysmal rupture in those patients. However, long-term follow-up is still required. Two patients without pulsation noted by 4D-CTA voluntarily requested neurosurgical clipping for prevention therapy due to concerns about their own severe hypertension. Intraoperative findings showed more stable homogenous walls (as shown in Fig. 8). To summarize, using 4D-CTA images, we found that irregular pulsation was closely associated with aneurysm rupture, and the discovery of the pulsation with unruptured aneurysms has important implications for the guiding treatment of the aneurysms. This finding is in line with those of some studies that reported irregular pulsation was correlated with the status of aneurysm rupture [10, 38]. Kato et al. observed that aneurysms with irregular pulsation had varied thicknesses and an absence of the internal elastic lamina and tunica media [17]. Krings et al. demonstrated that the presence of abnormal pulsation was related to aneurysm growth [39]. Similarly, Ferrari and colleagues reported that the irregular pulsation detected by 4D-CTA may potentially lead to a morphologic change in shape at follow-up and had a significant correlation with dark reddish thinner walls [40]. Therefore, it indicates that this method may be useful for identifying aneurysms with a higher risk of rupture.

Our study has some limitations: first of all, there is a lack of comparison with the results of surgery in terms of morphology, as a previous study showed a comparison with DSA could be sufficient to evaluate the morphology of aneurysms [41]. Secondly, the occurrence of thrombosis in the aneurysm wasn’t evaluated due to the absence of thrombosis in all aneurysms in this study. However, it is critical for the treatment plan decision-making whether or not thrombosis is present. Finally, the relatively small sample size may limit the power of the analysis. Hence, future prospective studies with larger patient populations and longer follow-up period are needed to further explore the role of 4D-CTA in the visualization of aneurysmal pulsation.